Question

在我的tblAccount中，有3个索引。聚簇索引是“Date，Accnumber，CifNum”。而非群集是“ID1”和“ID2”。我使用内部自联接来检测需要确定为错误的内容，但是存在大量数据并且连接需要很长时间，导致性能降低。可以做些什么来避免这种情况？

我想检查row1 ID1与row2 ID2是否相同，这将是一个错误。

UPDATE t SET ErrorCode = 'Same value of ID1 ID2'
FROM tblAccount t WITH (NOLOCK)
INNER JOIN tblAccount t2 WITH (NOLOCK)
ON t.Date = t2.Date
AND t.CifNum = t2.CifNum
AND t.ID2 = t2.ID1
WHERE t.Date = '20171231'

下面是为同一个CifNum和Date检测到不同的ID1或ID2。

UPDATE t SET t.ErrorCode = (CASE WHEN t.ID1 <> t2.ID1 OR t.ID2 <> t2.ID2 THEN 'Different ID1 or ID2' ELSE NULL END)
FROM tblAccount t WITH (NOLOCK)
INNER JOIN tblAccount t2 WITH (NOLOCK)
ON t.Date = t2.Date 
AND t.CifNum = t2.CifNum
AND (t.ID1 <> t2.ID1 OR t.ID2 <> t2.ID2 )
WHERE t.Date = '20171231'

在这种情况下，小组是否可以通过以下方式获得帮助？

UPDATE t SET ErrorCode = 'Same value of ID1 ID2'
FROM tblAccount t WITH (NOLOCK) 
inner join (select cifnum, ID1, ID2, Date from tblAccount WHERE Date ='20171231' group by cifnum, ID1, ID2, Date) t2
ON t.Date = t2.Date
AND t.CifNum = t2.CifNum
AND t.ID2 = t2.ID1
WHERE t.Date = '20171231'

Answer 1

您可以使用它来更新相同的值

UPDATE t SET ErrorCode = 'Same value of ID1 ID2'
FROM tblAccount t 
WHERE EXISTS (
    SELECT * FROM tblAccount t2 
        WHERE t.Date = t2.Date
        AND t.CifNum = t2.CifNum
        AND t.ID2 = t2.ID1)
AND t.Date = '20171231'

大量数据的Update Self Join性能降低

1 个答案: